0

How does it possible to identify throughput in MLRun solution (I use MLRun 1.3.0 with Iguazio version 3.5.2)?

I am using MLRun real-time function 'nuclio-risk-sentiment' and I would like to see request/response time of MLRun.

It is easy to see e.g. view to the Memory, CPU, Network I/O usage (see the board in grafana 'private / Kubernetes Pods'):

Grafana

Do you know where to see the duration of the response/request (real throughput)?

JzD
  • 22
  • 11

1 Answers1

0

The Iguazio uses for monitoring standard technology stack Prometheus and Grafana, it means, it is possible to see performance of NGINX (web server in Iguazio).

Grafana dashboard see 'private / NGINX - Request Handling Performance' and view to the 'Total request handling time':

Total request handling time

JIST
  • 1,139
  • 2
  • 8
  • 30